EM Model Component | Sample/Data Source | Analysis |
---|---|---|
I. Family caregivers’ (FCG) experiences of Ecological Model (EM) model constructs (categories 1,3,4,5,6,8) | FCGs who self-identify as black, Latino, Asian, and white will be randomly selected (approximately 10% of families from each ethnic group; if theoretical saturation is not reached, interviews will continue until saturation is reached.) A draft semi-structured interview guide will be refined with the input of a hospital patient / family council | Qualitative content analysis of audiotaped, transcribed interviews conducted at the 6-month post-hospital discharge home visit. Trustworthiness [113] is enhanced by the team’s methodological expertise, random sampling, rigorous analytic approaches, member checking, and a detailed audit trail |
II. Nurse champion experiences of EM model constructs (categories 1,3,4,5,6,8) | The 6 champions (2 per setting) will be interviewed using a semi-structured interview guide | Qualitative content analysis of audiotaped, transcribed interviews at conclusion of intervention at each site |
III. Evaluation of measures for reliability and validity for the ethnic groups represented in the study (category 7) | 1) The internal consistency of the caregiver outcome measures as well as relationships with known correlates (e.g., educational level) will be evaluated in the first 20 Spanish-speaking respondents/ 2) Both content and face validity of each measure will be discussed in participant interviews outlined in Component I above | Cronbach’s alphas for caregiver outcomes will be assessed. The participants will be asked to assess their perceptions of measures and identify potential cultural gaps in measurement content in Component I above |
IV. Assessment of whether ethnic concordance moderates the relationship between treatment group and patient/ family outcomes (category 2) | The ethnicities of nurse champions and each family will be documented as concordant or not | Concordance will potentially be used as a covariate in analyses for specific aims 1, 2 |